[Nuke-users] OCIO Colorspace (nuke-default) clamping colors

[Mads Lund]

I can see that the OCIO Colorspace node, using the nuke-default config,
clamps colors between 0-1 while the regular colorpsace node doesn't? Is it
a bug or a feature?


set cut_paste_input [stack 0]
version 9.0 v7
BackdropNode {
 inputs 0
 name BackdropNode1
 tile_color 0x7171c600
 label OCIO
 note_font_size 42
 selected true
 xpos 119
 ypos -38
 bdheight 162
}
BackdropNode {
 inputs 0
 name BackdropNode2
 tile_color 0x8e388e00
 label Regular
 note_font_size 42
 selected true
 xpos 343
 ypos -37
 bdheight 162
}
push $cut_paste_input
Expression {
 expr0 -0.5+(x/width)*2
 expr1 (x/width)>0.5?0.5:0
 name Expression6
 selected true
 xpos 237
 ypos -90
}
set Ndcf74c00 [stack 0]
Colorspace {
 colorspace_out sRGB
 name Colorspace20
 selected true
 xpos 353
 ypos 43
}
Colorspace {
 colorspace_in sRGB
 name Colorspace21
 selected true
 xpos 353
 ypos 69
}
Expression {
 expr1 abs((y*2-(height/2))-(r*height))<4?1:0
 name Expression7
 selected true
 xpos 353
 ypos 95
}
push $Ndcf74c00
OCIOColorSpace {
 in_colorspace linear
 out_colorspace sRGB
 name OCIOColorSpace4
 selected true
 xpos 129
 ypos 42
}
OCIOColorSpace {
 in_colorspace sRGB
 out_colorspace linear
 name OCIOColorSpace5
 selected true
 xpos 129
 ypos 68
}
Expression {
 expr1 abs((y*2-(height/2))-(r*height))<4?1:0
 name Expression10
 selected true
 xpos 129
 ypos 94
}
___
Nuke-users mailing list
Nuke-users@support.thefoundry.co.uk, http://forums.thefoundry.co.uk/
http://support.thefoundry.co.uk/cgi-bin/mailman/listinfo/nuke-users

Re: [Nuke-users] OCIO Colorspace (nuke-default) clamping colors

[Michael Habenicht]

Well, it is how OCIO works and how the nuke-default is implemented. It is all 
defined through 1D lookup tables. And that means it is defined only in a 
limited range. In this case it is -0.125 - 1.125. So when you convert back from 
linear it gets clamped to this range.
The colorspace node on the other hand uses mathematical formulars which are 
defined for the whole range.

Best regards,
Michael

On November 1, 2015 11:57:58 AM CET, Mads Lund  wrote:
>I can see that the OCIO Colorspace node, using the nuke-default config,
>clamps colors between 0-1 while the regular colorpsace node doesn't? Is
>it
>a bug or a feature?
>
>
>set cut_paste_input [stack 0]
>version 9.0 v7
>BackdropNode {
> inputs 0
> name BackdropNode1
> tile_color 0x7171c600
> label OCIO
> note_font_size 42
> selected true
> xpos 119
> ypos -38
> bdheight 162
>}
>BackdropNode {
> inputs 0
> name BackdropNode2
> tile_color 0x8e388e00
> label Regular
> note_font_size 42
> selected true
> xpos 343
> ypos -37
> bdheight 162
>}
>push $cut_paste_input
>Expression {
> expr0 -0.5+(x/width)*2
> expr1 (x/width)>0.5?0.5:0
> name Expression6
> selected true
> xpos 237
> ypos -90
>}
>set Ndcf74c00 [stack 0]
>Colorspace {
> colorspace_out sRGB
> name Colorspace20
> selected true
> xpos 353
> ypos 43
>}
>Colorspace {
> colorspace_in sRGB
> name Colorspace21
> selected true
> xpos 353
> ypos 69
>}
>Expression {
> expr1 abs((y*2-(height/2))-(r*height))<4?1:0
> name Expression7
> selected true
> xpos 353
> ypos 95
>}
>push $Ndcf74c00
>OCIOColorSpace {
> in_colorspace linear
> out_colorspace sRGB
> name OCIOColorSpace4
> selected true
> xpos 129
> ypos 42
>}
>OCIOColorSpace {
> in_colorspace sRGB
> out_colorspace linear
> name OCIOColorSpace5
> selected true
> xpos 129
> ypos 68
>}
>Expression {
> expr1 abs((y*2-(height/2))-(r*height))<4?1:0
> name Expression10
> selected true
> xpos 129
> ypos 94
>}
>
>
>
>
>___
>Nuke-users mailing list
>Nuke-users@support.thefoundry.co.uk, http://forums.thefoundry.co.uk/
>http://support.thefoundry.co.uk/cgi-bin/mailman/listinfo/nuke-users

-- 
Sent from my Android tab with K-9 Mail. Please excuse my brevity.
___
Nuke-users mailing list
Nuke-users@support.thefoundry.co.uk, http://forums.thefoundry.co.uk/
http://support.thefoundry.co.uk/cgi-bin/mailman/listinfo/nuke-users

Re: [Nuke-users] Export user tracks to Axis from CameraTracker

[Michael Garrett]

That worked...thanks!

On 1 November 2015 at 16:56, Frank Rueter|OHUfx  wrote:

> I haven't tried but you should be able to do this on any resolved 3D point
> (green points).
>
>
> On 1/11/15 2:50 am, Michael Garrett wrote:
>
> Thank you for pointing out that menu in the 2d view. So can I do that over
> any point including a user track, meaning a 2d track I've fed in?
>
> On 30 October 2015 at 21:57, Frank Rueter|OHUfx  wrote:
>
>> Ah, yes, you have to do this through the 2D view, not the 3D view.
>>
>> On 31/10/15 2:46 pm, Howard Jones wrote:
>>
>> I think Michael is looking at the point cloud, not the solved tracker.
>> Is that right?  The tracker has the options Frank is talking about.
>>
>> Howard
>>
>> On 31 Oct 2015, at 1:15 a.m., Frank Rueter|OHUfx  wrote:
>>
>> Are we talking about the same thing?
>> When I camera track something here and solve the tracks, I can right
>> click on any solved 3D point and create objects on it:
>> 
>>
>>
>>
>> On 31/10/15 7:04 am, Michael Garrett wrote:
>>
>> There's no right click option to do that from the point cloud when the
>> Viewer is on the CameraTracker, but after the point cloud is exported from
>> the tracker, the point can be selected in vertex mode and an Axis can be
>> snapped to it. I notice that survey points are added to the exported point
>> cloud as black points. Which would be hard to read if the point cloud is
>> dark.
>>
>> So that's probably faster than what I was doing. Thanks Frank! But it
>> would be good if there was a menu option to do what you're saying.
>>
>>
>>
>> On 29 October 2015 at 21:41, Frank Rueter|OHUfx  wrote:
>>
>>> I haven't looked at it in a while but can't you just right click on the
>>> solved point in the viewer to create an Axis at it's location?
>>>
>>>
>>>
>>> On 30/10/15 4:13 am, Michael Garrett wrote:
>>>
>>> I've been using the CameraTracker more these days and I noticed that
>>> when I feed in 2D user tracks, after the track Nuke projects them into 3D
>>> space which is potentially useful for generating an Axis.
>>>
>>> In the user track tab, the xyz position is available like in the older
>>> ProjectionSolver node, however I couldn't immediately see a way to export
>>> an Axis based on that solved 3D position. Of course it could automated
>>> pythonically, but I was wondering if there is an existing way. For now, I'm
>>> copy pasting the values out but I need to paste it into a text editor to
>>> get rid of the xy values first.
>>>
>>> Thanks,
>>> Michael
>>>
>>>
>>> ___
>>> Nuke-users mailing listnuke-us...@support.thefoundry.co.uk, 
>>> http://forums.thefoundry.co.uk/http://support.thefoundry.co.uk/cgi-bin/mailman/listinfo/nuke-users
>>>
>>>
>>> --
>>> [image: ohufxLogo 50x50]  *vfx compositing
>>>  | workflow customisation and
>>> consulting  *
>>>
>>> ___
>>> Nuke-users mailing list
>>> Nuke-users@support.thefoundry.co.uk, http://forums.thefoundry.co.uk/
>>> http://support.thefoundry.co.uk/cgi-bin/mailman/listinfo/nuke-users
>>>
>>
>>
>>
>> ___
>> Nuke-users mailing listnuke-us...@support.thefoundry.co.uk, 
>> http://forums.thefoundry.co.uk/http://support.thefoundry.co.uk/cgi-bin/mailman/listinfo/nuke-users
>>
>>
>> --
>> [image: ohufxLogo 50x50]  *vfx compositing
>>  | workflow customisation and
>> consulting  *
>>
>> ___
>> Nuke-users mailing list
>> Nuke-users@support.thefoundry.co.uk, http://forums.thefoundry.co.uk/
>> http://support.thefoundry.co.uk/cgi-bin/mailman/listinfo/nuke-users
>>
>>
>>
>> ___
>> Nuke-users mailing listnuke-us...@support.thefoundry.co.uk, 
>> http://forums.thefoundry.co.uk/http://support.thefoundry.co.uk/cgi-bin/mailman/listinfo/nuke-users
>>
>>
>> --
>> [image: ohufxLogo 50x50]  *vfx compositing
>>  | workflow customisation and
>> consulting  *
>>
>> ___
>> Nuke-users mailing list
>> Nuke-users@support.thefoundry.co.uk, http://forums.thefoundry.co.uk/
>> http://support.thefoundry.co.uk/cgi-bin/mailman/listinfo/nuke-users
>>
>
>
>
> ___
> Nuke-users mailing listnuke-us...@support.thefoundry.co.uk, 
> http://forums.thefoundry.co.uk/http://support.thefoundry.co.uk/cgi-bin/mailman/listinfo/nuke-users
>
>
> --
> [image: ohufxLogo 50x50]  *vfx compositing
>  | workflow customisation and
> consulting 

Re: [Nuke-users] OCIO Colorspace (nuke-default) clamping colors

[Rangi Sutton]

Hi Mads,

That's a feature, IMO.

sRGB doesn't define super brights. It should clamp.

I think you should get the same result saving an image out as sRGB, and
reading it back in, as converting it to sRGB and converting it back. Which
is what you see with the OCIO node.

Don't convert to a low-dynamic range colour space in the middle of your
processing tree!

r.


Rangi Sutton | VFX Supervisor
*CUTTING*EDGE  |  Brisbane Australia
Studio +61 7 3013 6236 | Mobile +61 41 521 1280

*/SYD /BNE /MEL /TYO*

On 1 November 2015 at 20:57, Mads Lund  wrote:

> I can see that the OCIO Colorspace node, using the nuke-default config,
> clamps colors between 0-1 while the regular colorpsace node doesn't? Is it
> a bug or a feature?
>
>
> set cut_paste_input [stack 0]
> version 9.0 v7
> BackdropNode {
>  inputs 0
>  name BackdropNode1
>  tile_color 0x7171c600
>  label OCIO
>  note_font_size 42
>  selected true
>  xpos 119
>  ypos -38
>  bdheight 162
> }
> BackdropNode {
>  inputs 0
>  name BackdropNode2
>  tile_color 0x8e388e00
>  label Regular
>  note_font_size 42
>  selected true
>  xpos 343
>  ypos -37
>  bdheight 162
> }
> push $cut_paste_input
> Expression {
>  expr0 -0.5+(x/width)*2
>  expr1 (x/width)>0.5?0.5:0
>  name Expression6
>  selected true
>  xpos 237
>  ypos -90
> }
> set Ndcf74c00 [stack 0]
> Colorspace {
>  colorspace_out sRGB
>  name Colorspace20
>  selected true
>  xpos 353
>  ypos 43
> }
> Colorspace {
>  colorspace_in sRGB
>  name Colorspace21
>  selected true
>  xpos 353
>  ypos 69
> }
> Expression {
>  expr1 abs((y*2-(height/2))-(r*height))<4?1:0
>  name Expression7
>  selected true
>  xpos 353
>  ypos 95
> }
> push $Ndcf74c00
> OCIOColorSpace {
>  in_colorspace linear
>  out_colorspace sRGB
>  name OCIOColorSpace4
>  selected true
>  xpos 129
>  ypos 42
> }
> OCIOColorSpace {
>  in_colorspace sRGB
>  out_colorspace linear
>  name OCIOColorSpace5
>  selected true
>  xpos 129
>  ypos 68
> }
> Expression {
>  expr1 abs((y*2-(height/2))-(r*height))<4?1:0
>  name Expression10
>  selected true
>  xpos 129
>  ypos 94
> }
>
>
> ___
> Nuke-users mailing list
> Nuke-users@support.thefoundry.co.uk, http://forums.thefoundry.co.uk/
> http://support.thefoundry.co.uk/cgi-bin/mailman/listinfo/nuke-users
>
___
Nuke-users mailing list
Nuke-users@support.thefoundry.co.uk, http://forums.thefoundry.co.uk/
http://support.thefoundry.co.uk/cgi-bin/mailman/listinfo/nuke-users

Re: [Nuke-users] Export user tracks to Axis from CameraTracker

[Frank Rueter|OHUfx]

I haven't tried but you should be able to do this on any resolved 3D 
point (green points).


On 1/11/15 2:50 am, Michael Garrett wrote:
Thank you for pointing out that menu in the 2d view. So can I do that 
over any point including a user track, meaning a 2d track I've fed in?


On 30 October 2015 at 21:57, Frank Rueter|OHUfx > wrote:


Ah, yes, you have to do this through the 2D view, not the 3D view.

On 31/10/15 2:46 pm, Howard Jones wrote:

I think Michael is looking at the point cloud, not the solved
tracker.
Is that right?  The tracker has the options Frank is talking about.

Howard

On 31 Oct 2015, at 1:15 a.m., Frank Rueter|OHUfx > wrote:


Are we talking about the same thing?
When I camera track something here and solve the tracks, I can
right click on any solved 3D point and create objects on it:




On 31/10/15 7:04 am, Michael Garrett wrote:

There's no right click option to do that from the point cloud
when the Viewer is on the CameraTracker, but after the point
cloud is exported from the tracker, the point can be selected
in vertex mode and an Axis can be snapped to it. I notice that
survey points are added to the exported point cloud as black
points. Which would be hard to read if the point cloud is dark.

So that's probably faster than what I was doing. Thanks Frank!
But it would be good if there was a menu option to do what
you're saying.



On 29 October 2015 at 21:41, Frank Rueter|OHUfx
> wrote:

I haven't looked at it in a while but can't you just right
click on the solved point in the viewer to create an Axis
at it's location?



On 30/10/15 4:13 am, Michael Garrett wrote:

I've been using the CameraTracker more these days and I
noticed that when I feed in 2D user tracks, after the
track Nuke projects them into 3D space which is
potentially useful for generating an Axis.

In the user track tab, the xyz position is available like
in the older ProjectionSolver node, however I couldn't
immediately see a way to export an Axis based on that
solved 3D position. Of course it could automated
pythonically, but I was wondering if there is an existing
way. For now, I'm copy pasting the values out but I need
to paste it into a text editor to get rid of the xy values
first.

Thanks,
Michael


___
Nuke-users mailing list
Nuke-users@support.thefoundry.co.uk

,http://forums.thefoundry.co.uk/
http://support.thefoundry.co.uk/cgi-bin/mailman/listinfo/nuke-users


-- 
ohufxLogo 50x50  	*vfx compositing

 | *workflow
customisation and consulting
* *


___
Nuke-users mailing list
Nuke-users@support.thefoundry.co.uk
,
http://forums.thefoundry.co.uk/
http://support.thefoundry.co.uk/cgi-bin/mailman/listinfo/nuke-users




___
Nuke-users mailing list
Nuke-users@support.thefoundry.co.uk
,http://forums.thefoundry.co.uk/
http://support.thefoundry.co.uk/cgi-bin/mailman/listinfo/nuke-users


-- 
ohufxLogo 50x50  	*vfx compositing

 | *workflow
customisation and consulting
* *

___
Nuke-users mailing list
Nuke-users@support.thefoundry.co.uk
,
http://forums.thefoundry.co.uk/
http://support.thefoundry.co.uk/cgi-bin/mailman/listinfo/nuke-users



___
Nuke-users mailing list
Nuke-users@support.thefoundry.co.uk
,http://forums.thefoundry.co.uk/
http://support.thefoundry.co.uk/cgi-bin/mailman/listinfo/nuke-users


-- 
ohufxLogo 50x50  	*vfx compositing

 | *workflow
customisation and consulting
* *


___
Nuke-users mailing list
Nuke-users@support.thefoundry.co.uk
,
http://forums.thefoundry.co.uk/
http://support.thefoundry.co.uk/cgi-bin/mailman/listinfo/nuke-users

Re: [Nuke-users] OCIO Colorspace (nuke-default) clamping colors

[Simon Björk]

Well, there's definetly cases were you want to a gamma encoded format in
your processing tree (for example sRGB). For example it will sometimes give
better result when keying by converting to sRGB first (and then back to
linear).

For this very reason OpenColorIO added support for a secondary sRGB lut
which will hold values up to around 5. It's called sRGBf (sRGB float). By
looking at the ocio folder that's included with Nuke I see that the lut is
actually there, but it's not added to the config. Have a look at the
official config

.

Or just add it back in by adding this to your config file:

- ! name: sRGBf family: "" equalitygroup: "" bitdepth: 32f
description: | Standard RGB Display Space, but with additional range to
preserve float highlights. isdata: false allocation: uniform
allocationvars: [-0.125, 4.875] to_reference: ! {src:
srgbf.spi1d, interpolation: linear}




---
Simon Björk
Compositor/TD

+46 (0)70-2859503
www.bjorkvisuals.com

2015-11-01 22:06 GMT+01:00 Rangi Sutton :

> Hi Mads,
>
> That's a feature, IMO.
>
> sRGB doesn't define super brights. It should clamp.
>
> I think you should get the same result saving an image out as sRGB, and
> reading it back in, as converting it to sRGB and converting it back. Which
> is what you see with the OCIO node.
>
> Don't convert to a low-dynamic range colour space in the middle of your
> processing tree!
>
> r.
>
>
> Rangi Sutton | VFX Supervisor
> *CUTTING*EDGE  |  Brisbane Australia
> Studio +61 7 3013 6236 | Mobile +61 41 521 1280
>
> */SYD /BNE /MEL /TYO*
>
> On 1 November 2015 at 20:57, Mads Lund  wrote:
>
>> I can see that the OCIO Colorspace node, using the nuke-default config,
>> clamps colors between 0-1 while the regular colorpsace node doesn't? Is it
>> a bug or a feature?
>>
>>
>> set cut_paste_input [stack 0]
>> version 9.0 v7
>> BackdropNode {
>>  inputs 0
>>  name BackdropNode1
>>  tile_color 0x7171c600
>>  label OCIO
>>  note_font_size 42
>>  selected true
>>  xpos 119
>>  ypos -38
>>  bdheight 162
>> }
>> BackdropNode {
>>  inputs 0
>>  name BackdropNode2
>>  tile_color 0x8e388e00
>>  label Regular
>>  note_font_size 42
>>  selected true
>>  xpos 343
>>  ypos -37
>>  bdheight 162
>> }
>> push $cut_paste_input
>> Expression {
>>  expr0 -0.5+(x/width)*2
>>  expr1 (x/width)>0.5?0.5:0
>>  name Expression6
>>  selected true
>>  xpos 237
>>  ypos -90
>> }
>> set Ndcf74c00 [stack 0]
>> Colorspace {
>>  colorspace_out sRGB
>>  name Colorspace20
>>  selected true
>>  xpos 353
>>  ypos 43
>> }
>> Colorspace {
>>  colorspace_in sRGB
>>  name Colorspace21
>>  selected true
>>  xpos 353
>>  ypos 69
>> }
>> Expression {
>>  expr1 abs((y*2-(height/2))-(r*height))<4?1:0
>>  name Expression7
>>  selected true
>>  xpos 353
>>  ypos 95
>> }
>> push $Ndcf74c00
>> OCIOColorSpace {
>>  in_colorspace linear
>>  out_colorspace sRGB
>>  name OCIOColorSpace4
>>  selected true
>>  xpos 129
>>  ypos 42
>> }
>> OCIOColorSpace {
>>  in_colorspace sRGB
>>  out_colorspace linear
>>  name OCIOColorSpace5
>>  selected true
>>  xpos 129
>>  ypos 68
>> }
>> Expression {
>>  expr1 abs((y*2-(height/2))-(r*height))<4?1:0
>>  name Expression10
>>  selected true
>>  xpos 129
>>  ypos 94
>> }
>>
>>
>> ___
>> Nuke-users mailing list
>> Nuke-users@support.thefoundry.co.uk, http://forums.thefoundry.co.uk/
>> http://support.thefoundry.co.uk/cgi-bin/mailman/listinfo/nuke-users
>>
>
>
> ___
> Nuke-users mailing list
> Nuke-users@support.thefoundry.co.uk, http://forums.thefoundry.co.uk/
> http://support.thefoundry.co.uk/cgi-bin/mailman/listinfo/nuke-users
>
___
Nuke-users mailing list
Nuke-users@support.thefoundry.co.uk, http://forums.thefoundry.co.uk/
http://support.thefoundry.co.uk/cgi-bin/mailman/listinfo/nuke-users